S S symmetry
Article Composite and Background Fields in Non-Abelian Gauge Models
Pavel Yu. Moshin 1,* and Alexander A. Reshetnyak 1,2,3
1 Department of Physics, National Research Tomsk State University, 634050 Tomsk, Russia; [email protected] 2 Center for Theoretical Physics, Tomsk State Pedagogical University, 634061 Tomsk, Russia 3 Institute of Strength Physics and Materials Science Siberian Branch of Russian Academy of Sciences, 634021 Tomsk, Russia * Correspondence: [email protected]
Received: 19 October 2020; Accepted: 25 November 2020; Published: 30 November 2020
Abstract: A joint introduction of composite and background fields into non-Abelian quantum gauge theories is suggested based on the symmetries of the generating functional of Green’s functions, with the systematic analysis focused on quantum Yang–Mills theories, including the properties of the generating functional of vertex Green’s functions (effective action). For the effective action in such theories, gauge dependence is found in terms of a nilpotent operator with composite and background fields, and on-shell independence from gauge fixing is established. The basic concept of a joint introduction of composite and background fields into non-Abelian gauge theories is extended to the Volovich–Katanaev model of two-dimensional gravity with dynamical torsion, as well as to the Gribov–Zwanziger theory.
Keywords: composite fields; background fields; Yang–Mills theory; Volovich–Katanaev model; Gribov–Zwanziger model; background effective action; gauge-dependence
1. Introduction Composite [1,2] and background [3–5] fields are widely used in quantum gauge theories. The attention to composite fields (see [2] for an overview) stems from the fact that the effective action for composite fields suggested in [1] has been applied to quantum field models such as [6–8], including the Early Universe, Inflationary Universe, Standard Model and SUSY theories [9–13]. The study of BRST-invariant renormalizability in Yang–Mills theories, which includes N = 1 SUSY formulations [14–16], the functional renormalization group [17–21] and the Gribov horizon [22–24], appears to be promising within the concept of soft BRST symmetry breaking [25–28] and the local composite operator technique [29,30] as applied to arbitrary backgrounds [31]. In its turn, the background field method [3–5] presents the quantization of Yang–Mills theories using background gauges [5,32,33] in such a way that provides an invariance of the effective action under the gauge transformations of background fields and reproduces physical results with essential simplifications in the Feynman diagrams, thereby providing insight into diverse quantum properties of gauge theories [34–43]; for recent developments, see [44–48]. The present article is devoted to quantum non-Abelian gauge models with composite and background fields, whose consistent analysis isfocused on Yang–Mills theories quantized using the Faddeev–Popov method [49] in a combined presence of composite and background fields. A joint treatment of Yang–Mills fields Aµ with composite and background ones requires a systematic consideration of these ingredients as interrelated. We forward the symmetry principle as such a systematic concept. In fact, suppose that a generating functional Z (J, L) of Green’s functions with
Symmetry 2020, 12, 1985; doi:10.3390/sym12121985 www.mdpi.com/journal/symmetry Symmetry 2020, 12, 1985 2 of 26
A composite fields is given, depending on sources JA for the quantum fields φ , as well as on sources m Lm for the composite fields σ (φ). A question then naturally arises of how one can introduce some background fields Bµ in a way that leads to an extended functional Z (B, J, L) reflecting the possible symmetries of Z (J, L). Let us next suppose that a generating functional Z (B, J) of Green’s functions, also with certain symmetries, is given in the background field method, and then the question is how m some composite fields σ (φ, B) with sources Lm can be introduced for the resulting Z (B, J, L) to inherit the given symmetries. It turns out that these two approaches are equivalent in the following sense. According to the first approach, a generating functional Z (J, L) is given,
Z i h i Z (J, L) = dφ exp S (φ) + J φA + L σm (φ) , (1) h¯ FP A m corresponding to the Faddeev–Popov action SFP (φ) of a Yang–Mills theory with composite fields m σ (φ). A background field Bµ can then be introduced by localizing the inherent global symmetry of Z (J, L) under SU(N) transformations (rotations for JA and tensor transformations for Lm) in such a way that Z (B, J, L) defined as Z (B, J, L) = Z (J, L)| (2) ∂µ→Dµ(B) is invariant under local SU(N) transformations of the sources JA, Lm, accompanied by gauge transformations of the field Bµ with an associated covariant derivative Dµ (B), where the precise meaning of ∂µ → Dµ (B) in (2) is given by (19) of Section2. The original SFP (φ) becomes thereby modified to the Faddeev–Popov action of the background field method, SFP (φ, B), which is related to SFP (φ) by so-called background and quantum transformations of this method (see Section 2.2). According to the second approach, a generating functional Z (B, J) is constructed using the background field method for Yang–Mills theories,
Z i h i Z (B, J) = dφ exp S (φ, B) + J φA , (3) h¯ FP A which implies S (φ, B) = S (φ)| . FP FP ∂µ→Dµ(B) m Some composite fields σ (φ, B) with sources Lm can then be introduced on condition that the resulting generating functional
Z i h i Z (B, J, L) = dφ exp S (φ, B) + J φA + L σm (φ, B) (4) h¯ FP A m should inherit the symmetry of Z (B, J) under local SU(N) rotations of the sources JA accompanied by gauge transformations of the background field Bµ with the covariant derivative Dµ (B). This symmetry requirement for Z (B, J, L) is satisfied by a local SU(N) tensor transformation law imposed on σm (φ, B) m and is provided by Bµ entering the composite fields σ (φ, B) by means of the covariant derivative Dµ (B), which implies σm (φ, B) = σm (φ)| (5) ∂µ→Dµ(B) for certain σm (φ), and thereby we return to the first approach. In the main part of the article, we implement the first approach as a starting point of our systematic analysis, assuming the composite fields to be local, whereas in the remaining part we show how the first and second approaches can be extended beyond the given assumptions by considering the Volovich–Katanaev model of two-dimensional gravity with dynamical torsion [50] and the Gribov–Zwanziger theory [23,24]. Two-dimensional models of gravity [50–61] and supergravity [62–65] are of interest in view of their close relation to string and superstring theory. Simple two-dimensional models also provide a deeper insight into classical and quantum properties of gravity in higher dimensions, while in some Symmetry 2020, 12, 1985 3 of 26 cases these models are exactly solvable at the classical level. One of the two-dimensional gravity models that has been widely discussed at the classical [66–70] and quantum [71–75] levels is the model [50] suggested in the context of bosonic string theory with dynamical torsion [76] in order to address some problems of string theory. Thus, it has been shown [76], using the path integral approach, that a string with dynamical torsion has no critical dimension. The model [50] presents the most general theory of two-dimensional R2-gravity with independent dynamical torsion leading to second-order equations of motion for the zweibein and Lorentz connection. The model also contains solutions with constant curvature and zero torsion, thereby incorporating some other two-dimensional gravity models [51,52,61] whose actions, as compared to that of [50], do not allow a purely geometric interpretation. Being quantized in the background field method, the model yields a gauge-invariant background effective action [75]. In quantum Yang–Mills theories using differential (e.g., Landau or Feynman) gauges, the non-Abelian nature of the gauge group features the Gribov ambiguity [22] implying a residual gauge-invariance due to Gribov copies, which are removed by means of a Gribov horizon [22] implemented in the Gribov–Zwanziger model [23,24] being a quantum Yang–Mills theory in Landau gauge and including an additive horizon functional in terms of a non-local composite field [77]. The issue of bringing the horizon functional to other gauges has been settled due to the concept of finite field-dependent BRST transformations [26,28,78–84], which allows one to present the horizon functional using different gauges in a way consistent with the gauge-independence of the path integral, based on the Gribov–Zwanziger recipe [23,24] and starting from a BRST-invariant Yang–Mills quantum action in Landau gauge. The horizon functional in covariant Rξ gauges has been given by [26,28,82,85] (see also [86,87] for a BRST-invariant horizon) and later extended to the Standard Model in [88,89]. The article is organized as follows. In Section2, a generating functional of Green’s functions with composite and background fields in Yang–Mills theories is introduced, as well as a generating functional of vertex Green’s functions (effective action). Analyzing the dependence of the generating functionals of Green’s functions upon a choice of gauge-fixing, we find a gauge variation of the effective action in terms of a nilpotent operator depending on the composite fields and determine the conditions of on-shell gauge-independence. Besides, the effective action Γeff (B, Σ), with a background field Bµ and a set of auxiliary tensor fields Σm associated with σm, is found to exhibit a local symmetry m under the gauge transformations of Bµ combined with the local SU(N) transformations of Σ . In Section3, we examine the Volovich–Katanaev model [ 50] quantized according to the background field method in [75]. As an extension of our second approach (3)–(5) beyond the Yang–Mills case, the quantized two-dimensional gravity [75] is modified by the presence of local composite fields, and the corresponding background effective action is found to be gauge-invariant in a way similar to the Yang–Mills case. In Section4, we modify the Gribov–Zwanziger model [ 23,24] by introducing a background field, which extends our first approach (1), (2) beyond the Yang–Mills case with local composite fields and thereby yields a gauge-invariant background effective action. Section5 is devoted to concluding remarks. We use DeWitt’s condensed notation [90]. The Grassmann parity and ghost number of a quantity F are denoted by e (F), gh (F), respectively. The supercommutator [F, G} of any quantities F, G with definite Grassmann parities is given by [F , G} = FG − (−1)e(F)e(G)GF. Unless specified by an arrow, derivatives with respect to fields and sources are regarded as left-hand ones.
2. Generating Functionals and Their Properties Let us examine a generating functional Z (J, L) corresponding to the Faddeev–Popov action SFP (φ) of a Yang–Mills theory with local composite fields,
Z i h i Z (J, L) = dφ exp S (φ) + J φA + L σm (φ) , (6) h¯ FP A m Symmetry 2020, 12, 1985 4 of 26
m where Lm are sources to the composite fields σ (φ),
1 σm (φ) = Λm φAn ... φA1 , (7) ∑ A1...An n=2 n!
A i α α α i and JA are sources to the fields φ = A , b , c¯ , c composed by gauge fields A , (anti)ghost fields c¯α, cα, and Nakanishi–Lautrup fields bα, with the following distribution of Grassmann parity and ghost number:
A A A m A m e(φ ) = (0, 0, 1, 1) , gh(φ ) = (0, 0, −1, 1) , e(JA, Lm) = e(φ , σ ), gh(JA, Lm) = −gh(φ , σ ).
The Faddeev–Popov action SFP (φ), ←− SFP (φ) = S0 (A) + Ψ (φ) s , is given in terms of a gauge-invariant classical action S0 (A), invariant, δξ S0 (A) = 0, i i α under infinitesimal gauge transformations δξ A = Rα (A) ξ with a closed algebra of gauge i generators Rα (A), ←− i j i j γ i γ i i δ R (A) R (A) − R (A) Rα (A) = F R (A) , F = const, R ≡ R , α, j β β, j αβ γ αβ α, j α δAj ←− and a nilpotent Slavnov variation s applied to a gauge Fermion Ψ (φ), e (Ψ) = 1,
←− α ←− 2 SFP (φ) = S0 (A) + Ψ (φ) s , Ψ (φ) = c¯ χα (φ) , s = 0, (8) where A←− i α α α γ β φ s = Rα (A) c , 0, b , 1/2Fβγc c . For the explicit field content
i = (x, p, µ), α = (x, p) , µ = 0, . . . , D − 1, p = 1, . . . , N2 − 1, φA = Ap|µ, bp, c¯p, cp , (Aµ, b, c¯, c) ≡ Tp Ap|µ, bp, c¯p, cp , [Tp, Tq] = f pqrTr,
←− the field variations φA s have the form